INDEX

Explanations

programming-related keywords and constructs

oai_token-act-pair · gpt-4o-mini Triggered by @bot

Code, paths, and unconventional punctuation

np_acts-logits-general · gemini-2.0-flash

path, location, treatment

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

checkFor

-0.56

<<"\

-0.54

maxcdn

-0.53

MenuInflater

-0.52

ׁ

-0.51

checkIf

-0.51

("")]

-0.49

 iVar

-0.48

ieteur

-0.47

gatsby

-0.47

POSITIVE LOGITS

Ⅰ

0.72

NameInMap

0.69

 sortes

0.69

 chaude

0.69

 الرياضيه

0.68

 Abnormal

0.68

 coopération

0.67

saraba

0.66

Ⅱ

0.66

 matchs

0.66

Activations Density 0.388%

programming-related keywords and constructs

Code, paths, and unconventional punctuation

path, location, treatment

No Comments

No Known Activations